60 days--NIST Refuses To Reply

It has been 60 days since I asked NIST about the numerous discrepancies between Frankel Fabrication Shop Drawing #9114 and certain figures found in their WTC7 report NCSTAR 1-9.

Frankel Drawing #9114 clearly shows stiffener plates welded onto the end of the girder that NIST claims walked off its C79 seated connection on floor 13, leading to an unprecedented global progressive collapse.

Please contact NIST Director Patrick Gallagher and Public Affairs Officer Michael Newman at the following email addresses and ask for a prompt reply to these pertinent questions.


You may phone them at:

Patrick Gallagher-- 301-975-2300 (Director )
Michael Newman-- 301-975-3025 (Public Affairs)


To: michael.newman@nist.gov; patrick.gallagher@nist.gov
Subject: WTC7 Report Discrepancies
Date: Tue, 24 Sep 2013 23:31:55 -0400

Dear Mr. Newman and Mr. Gallagher,

It has been 60 days since I sent my inquiry regarding the figures in NCSTAR 1-9.

I copy the original emails below for you convenience.

When may I expect a reply?

Thank you very much,
David Cole


To: michael.newman@nist.gov; patrick.gallagher@nist.gov
Subject: WTC7 Report Discrepancies
Date: Fri, 26 Jul 2013 14:48:48 -0400

Dear Mr. Newman,

On March 16, 2012 I wrote your office with an inquiry about certain errors in the NCSTAR 1-9 report. Your reply follows:

From: michael.newman@nist.gov
Date: Tue, 20 Mar 2012 16:49:50 -0400
Subject: Your Inquiry on NIST NCSTAR 1-9

Dear Mr. Cole,

Joseph Main forwarded your e-mail of March 16 to me for handling. Our researchers looked into the issue you raised and found that there is indeed an error in the drawing number cited for Figure 8-16 in NIST NCSTAR 1-9, Structural Fire Response and Probable Collapse Sequence of World Trade Center Building 7. The drawing used in Figure 8-16 was “Structural Drawing S8” rather than “Erection Drawing E12/13”. Figure 8-16 is used to locate features of the floor framing, and either drawing would serve this purpose. The differences between the two drawings are minor, involving some notes that appear in one drawing but not the other. The information of interest is the same in both drawings, and the error in the drawing number referenced does not affect any observations, findings or conclusions in the NCSTAR 1-9 report.

NIST will include an erratum to the report to indicate this correction.

Thank you for catching this error.
Michael Newman
NIST Public Affairs

********************************************************************* Given that as a result of my inquiry, NIST issued an Erratum (April 2012) concerning Figure 8-16, making reference to a specific drawing (structural drawing, Cantor S-8), and given the fact that a similar Erratum (June 2012) addressed the seat width and walk off distances for the 13th floor, column 79 girder connection, also made reference to a specific drawing (Frankel #1091), I now have a number of pertinent questions regarding other Figures in NCSTAR 1-9.

Technical Questions:

In NCSTAR 1-9, which design drawing was used to create

• Figure 8-21?
• Figure 8-23?
• Figure 8-26?
• Figure 8-27?
• Figure 11-16?
• Figure 11-19?
• Figure 12-24?
• Figure 12-25?

Given that Frankel drawing #9114 shows 3/4" web/flange stiffeners installed on the girder at the 13th floor column 79 connection, why weren't the stiffeners reported in NCSTAR 1-9 and shown in the figures listed above?

Was Frankel Drawing #9114 used? If not, why not?

Thank you very much,
David Cole


... will watch this with interest.


Can you post or publish a link to some kind of visual that makes the point about the "web/flange stiffeners" in a more graphic form? I've spent about an hour looking at the report and looking at the diagram 9114 and I can't identify any structure that would restrain the girder from moving.

Girder Movement

The stiffeners strengthen the girder end, transferring the load to the 2" thick underseat plate through the seat plate. They make the footprint of the girder wider, otherwise the load would be concentrated at or near the junction of web and flange.

They do not restrain the girder from moving. What they do is prevent the flange from folding in the unlikely scenario that the girder could move westerly across the seat, which eventually means the web would reach the edge of the seat. If this could happen (very unlikely) the flange would fold without the stiffeners in place.

I say very unlikely because there are a number of strange conditions offered by NIST that are necessary for the girder to even begin to move:

1. 4 hours of fire
2. Shear studs in the composite floor system break throughout the region allowing the beams to expand
3. Beams to the east thermally expand breaking the bolts holding the girder to the seat
4. No studs installed on the girder
5. Beams somehow expand 6.25"

However, if the beams expand, then so does the girder, which would trap the girder between the side plates of C79, limiting any movement to about 3.5".

NIST initially said the movement was 5.5", but changed this to 6.25" after we notified them that the seat was actually 12" wide, not 11". So you can see that 5.5" is impossible when girder expansion is understood.

See the evidence here: http://911blogger.com/news/2013-11-06/case-against-nist

The graphics there show a hypothetical 8.75" displacement (no heating), where red is high stress and blue is low. This proves that even at 2.5" further than NIST's distance correction of 6.25" the flange did not fold and the girder stays on the seat.

Result: No girder failure to initiate a global progressive collapse.

Here is a labeled displacement model for terminology.


Girder Movement

So, if I read you right, and thanks for the diagram, this has nothing to do with whether or not the girder could be moved enough to then be rocked off the seat by the beams. If that's correct, then I'm not sure I see why this is important to any discussion of the NIST simulation. Since the simulation didn't show the flanges folding and you find that unlikely in any case, why is it relevant to the NIST's explanation of the collapse mechanism? I do understand your other points, just this one baffles me as to relevance. You seem to actually want to make the point that the girder NEEDS to fold which both you and the NIST agree didn't happen, in order for the girder to move off the seat but I still can't (sadly) understand your argument here.


Who mentioned rocking off? You asked if the stiffeners prevent the girder from moving. I explained that.

You clearly understand more about the connection.

You now mention the NIST simulation. Why not tell us what you see in the simulation that doesn't comport with the narrative. Yeah, it's confusing. I suggest you scrutinize the sim and the narration and you may come a deeper understanding.

It's the flange that NIST says failed so the girder could fall from the seat. I say this is impossible because the stiffeners prevent flange failure.


Two follow ups. First, can you be precise about what flange you are saying that the NIST simulation showed as failing? I'm thinking that it's the bottom flange on the girder which is being pushed west but I'm not certain that's what you're saying. Second, where does it say in the report that this flange "failed to the girder could fall from the seat"? Every time I look at the document what I see is that the girder is supposed to be moving westward by several inches but I can't find anything in the report that says that the flange failed and that this failure caused the girder to fall off it's seat. Of course I can see that once the girder is only partially supported at some point there is a lot of force pressing on that flange but I don't see where flange failure is called out in the report. To be clear YOU didn't mention "rocking off" - that was my understanding of how the girder gets off it's seat. I'm not sure what you mean by "the narrative" so I can't respond to that. I'd prefer to keep the focus on your comments about the importance of the web stiffeners and your point about the failure of the girder flange.

NIST's words

On Page 488 (PDF page 554) of NCSTAR 1-9 see this:

Gravity shear loads in a beam were transferred to the bearing seat primarily in the proximity of the web on the bottom flange. Therefore, when the web was no longer supported by the bearing seat, the beam was assumed to have lost support, as the flexural stiffness of the bottom flange was assumed to be insufficient for transferring the gravity loads. Under such conditions, the beam was removed.

The only flange that NIST can be claiming failure on is the east flange.

Please show us the SIM you are referring to. I am not aware of any SIM showing the flange failure. Just narratives.

Here is the girder, pushed 8.75" to the west. With stiffeners, the east flange cannot fail. (No heating was applied to this FEA)


That´s it

"as the flexural stiffness of the bottom flange was assumed to be insufficient for transferring the gravity loads"

Compare this to NIST´s reply to why the stiffener plates were omitted: "The web stiffeners shown at the end of the girder in Frankel drawing #9114 prevent web crippling. The structural analyses of WTC 7 did not show any web crippling failures. Therefore, the web crippling plates did not need to be included in the models/analyses."

Peculiar* that they somehow did not realize that the model still assumes failure due to lack of stiffness at that very same place!

Not Exactly What I Asked

That quote simply describes how the simulation worked. (When I used the term simulation, I was referring generally to the computer simulations run by the NIST which led them to their conclusions about the most likely causes for the collapse of WTC7). It doesn't appear to represent any assertion that the web on the girder (between columns 44 and 79 as I recall) failed or that such a failure led to the global collapse of the structure. That's actually all I was asking. All it says is that WHEN the simulation showed the "web no longer supported" then the beam was removed from further inclusion in the simulation. So, I'm left, I think, with my original question. How does this have anything to do with the movement of the girder which is central to the collapse sequence? I thought that the key point about that girder is on page 536 which says that the girder to column connection at column 79 failed due to bolts shearing and to the girder being pushed off the bearing seat by thermal expansion. So, I still don't see that the presence or absence of those stiffeners is damaging to the credibility of the simulation.

past the half-way point..

I think I understand your question. Look at the video below from dtg86 for reference. NIST assumes that once the heat expansion causes the beam to move past the half-way point off the seat(those 6.25 inches or whatever), then it fails because the lower left corner of the beam connection(see video) is too soft to bear the load. With the stiffener plate this is impossible as the supported(stiffened) connection is not too soft and would have to be pushed the whole distance(another 6.25 inches), i.e. completely off the girder.


I looked at the video and I read your explanation, so now I'm certain that I understand your point. I'm sorry that I can't agree with it. The NIST report never says that the girder fails because the "lower left hand corner" (I think you must mean the girder) fails. That's not an unreasonable idea but that's not what's in the report as I read it. I hope I'm not confusing beams with girders, I'm not a structural engineer but my point is with regards to the girder which traverses the space between columns 79 and 44. All I actually know, again from my reading, is that the simulation shows that the girder moved enough so that it could no longer support any significant load.

It actually says..

NIST says the flange fails after being moved 6.25 inches "as the flexural stiffness of the bottom flange was assumed to be insufficient for transferring the gravity loads" - in other words it is assumed to be too soft on its own to carry the burden. The NIST report does not say "lower left hand corner", that is my explanation to you with reference to the video.

And if you watch the video again with this in mind, you realize that the assumed failure zone is the bottom left of that "bottom flange" NIST refers to("insufficient flexural stiffness").

But as gerrycan1 and others are pointing out, this lack of stiffness is due to the omission of the aptly named "stiffener plate", which invalidates the "assumption" by NIST.

It really is as simple as that.

Not Really So Simple

The video is your own creation and really cannot be cited in this manner but it really made the point very clear. The NIST report contains the wording you've cited but this refers to the criteria for removing beams from the analysis and not specifically to the collapse sequence. I can't read this the way you do, that it must mean that this particular girder suffered this failure. It's germane to note that, in the part of the report that details the collapse sequence, this is not mentioned as pertaining to the girder. Finally, this paragraph and another one with similar construction in the report refer to beams, not girders. I think that's why the NIST is saying that your thesis is irrelevant - this was the failure mechanism for some of the floor beams but not the girder that seems to be what you asked the NIST in the first place. So, the girder had a stiffener but the failure mechanism was different, the report says "walk off" and "bolt shear" not "insufficient flexural stiffness". That's how I read the report. Sorry that I can't agree with your analysis. I do think that they could have been clearer in the diagrams.

The "walk-off" in detail..

Listen, you say that the report says "walk off". More specifically NIST says 6.25 inches of "walking"...but the girder rests on a plate that is 12" wide...

so how come it falls off at only 6.25 inches?

Now stop the video at about 3.39 and see how this looks. The girder does not "walk" the whole way off the column, it goes a bit more than half-way (6.25 inches), at which point it is resting on one side of an upside-down "T", and it is this horizontal part of the upside-down "T" that fails due to lack off stiffness.

The video is not mine by the way, I just cited it.

How NIST defines "walk-off"

To be sure we should look at the term "walk-off" in its full context as NIST defines it.

Thermal Effects on Connections for Floor Beams and Girders

"Thermal expansion of beams and girders also caused connection failures. Restrained thermal expansion
of steel beams and girder within the structural system resulted in (1) bolt shear due to increased axial
forces, (2) walk-off of seated connections after bolts had sheared, and (3) failure of connection welds to
beam webs under shear forces.

Shear failure of all the bolts in fin and knife connections, or failure of the weld to the beam or girder web
in header connections, resulted in a loss of horizontal and vertical support to the beam or girder. In seated
connections (SWC, STP, and STC), the shear failure of bolts at the bearing seat and top clip or plate,
caused loss of horizontal support but not vertical support. As the east floor beams and the girder
continued to thermally expanded, the four bolts at the seated connection were sheared, resulting in a loss
of horizontal support at the connection. Loss of vertical support occurred when the beam or girder
“walked off” the bearing seat or when the bearing seat weld failed. Walk-off occurred when beams that
framed into the girders from one side thermally expanded and the resulting compressive forces in the
beams pushed laterally on the girder from one side, sheared the girder bolts, and then continued to
laterally push the girder until it walked off the bearing seat.

A girder was considered to have lost vertical support when its web was no longer supported by the
bearing seat. The bearing seat at Column 79 was 11 in. (Rev 12 in.) wide. Thus, when the girder end at Column 79
had been pushed laterally at least 5.5 in (Rev. 6,25)., it was no longer supported by the bearing seat.
factors that contributed to this failure were the absence of shear studs on the girders that would have
provided lateral restraint and the one-sided framing of the northeast corner floor beams that allowed the
floor beams to push laterally on the girder due to thermal expansion." -- NCSTAR 1-9 Vol 2 Pg 525 pdf (emphasis mine)


Let´s hope woodbourne finds your explanation better than mine.

To clarify, the "web" is the middle part of the girder, or the "I"? Did NIST back up with calculations the claim that the rest of the girder is too soft to bear the weight?

The only thing that bothers me is the alternative mode of failure, "when the bearing seat weld failed", which apparently is not assumed to have been the failure mode at the crucial Column 79. Did NIST give evidence for how and why the seat could have failed, and could it possibly offer this mode as an alternative for the walk off in case it accepts you have proven the walk off to be impossible?

Re walkoff...

I feel I need to chip in here because there is a question in my mind re the "walking off" (which may be what Woody is getting at). To explain, here is a simple analogy...

There is a drinks mat on my desk here in front of me. When I push it over the edge of the desk it gets to half way and falls off. I could say it tips off or rocks off. So, in the case of the girder - as it moves off the east side of its seat, when it gets over half way, it tips to the East (clockwise). In this scenario, the bottom, left flange has no bearing.

I hope this makes sense and look forward to your comments.


At some point the combination of gravity loading and walk off causes the girder to fall off it's seat. Exactly how this happens frankly I don't know. It's not the case that when you have a complex set of forces acting on a building which is losing it's structural integrity you can just can just claim that the only way it could have happened was if a web crippling failure (still hoping I'm using the term correctly) occurred within the structure of the girder. To respond to the post more directly, your analogy allows for only the force of gravity to be acting on your mat, what was going on in the collapse was vastly more complicated.

Walk vs Rock

You wrote: So, in the case of the girder - as it moves off the east side of its seat, when it gets over half way, it tips to the East (clockwise). In this scenario, the bottom, left flange has no bearing.

What you describe doesn't consider that the girder is pinched between the seat and the floor above over its entire length. Beams framing in the from the east keep it plumb.

Your drinks mat analogy is too flat and flexible. The girder is 33" tall, very strong.

The girder allegedly moves west off the west side of the seat. The east girder flange is the one that would necessarily fail--absent the stiffeners.

Here is the girder, with stiffeners in place, at 8.75" westerly displacement. The east flange in the foreground is not going to fail.


But look, you can't have just beam expansion due to fire. The girder will expand too and it will be trapped by the western side plate, limiting movement to ~3.5"

Be careful. Consider all conditions.

Thanks for the reply but...

Even if the girder is pinched, the floor is weighing down on it. If it gets too far off its seat, could the girder and the floor collapse?

I realise my mat analogy is simplistic (and I got east and west the wrong way round). It was just to demonstrate the movement rather than the whole system. It was an alternative walk off scenario where the stiffeners do not play a part.

For the record, I have major issues with the NIST initiation event...
- The expanding beams only seem to be expanding one way. Why are the not pushing at the other end?
- The girder that gets pushed off its seat. Would it not more likely bend? Or more likely, the heated expanding beams sag?
- Even if the girder does get pushed off its seat? I just don't buy that causing a global collapse.

I am just slightly concerned that too much emphasis is being put on this one particular issue - the stiffeners.

Beam Expansion

Yes, you bring up a good point. NIST does say that the beams expand west. There would be bolt failure at the east exterior columns and the 1" gap there would be taken up before the girder begins to push, if it could push.

All of these movements depend upon giving NIST several major concessions:

1. 4 hour fire, instead of 30 minutes
2. Shear stud failure before 300C
3. Exterior beam to column connections remain intact.
4. Girder doesn't expand enough to get trapped in the side plates
5. Beams can expand 6.25" (Damn!)

But even after all this, once the girder begins to move west it can go in excess of 8.75 inches and still maintain structural integrity, carrying the floor load above. The FEA proves this.

Here is a shot of the east C38 connection. If NIST has bolts breaking on both ends of the girder, bolts breaking on the beam to girder fin plates, why aren't the east exterior connections giving up also?


Your Issues

1. The expanding beams DO push both ways. It's just the case that the exterior column connections are stronger than the ones at the girder. So the girder moves and not the exterior columns. The real point is that the force of expansion breaks the bolt connections at a really low temperature.
2. The girder bends in the middle where you would expect it to since there is no steel member preventing it from doing so and no shear studs to anchor it to the flooring. The heated beams apparently first expand and then twist and sag, you are correct.
3. Read the sequence on pages 352 and 353 to get a better idea of the actual sequence shown by the simulation. It's more complicated, as you expected, than just the girder failing.

limited to 3.5"

I never noticed this before. How can NIST slither away from this one?

Also, is it not correct the the failure of the seat plate(pf) could not allow the girder to fall down since there is another plate(pg) directly below it, and since the girder is pressed up against the column it would still rest on pg?

No Problem

This is fine, but has nothing to do with whether or not the girder had a web crippling (hope I'm using the term correctly) failure. That was the question posed by the OP to the NIST and their answer agrees with your(?) video- no web crippling failure occurred. I just don't see your point at all here, sorry.


Yes, you can see that there is no mention of any form of web crippling failure here. That means that, unless you have some other citations, I'll stick to my premise which is that there was none mentioned in the report and thus the NIST's response to you makes perfect sense. The plates aren't in the diagrams because there presence didn't prevent walk off. That's because when they ran the simulation it didn't show that the girder failed in this manner. I do understand that this is just one facet of the concerns you have over that report, and honestly, I find the report more obscure than I'd like it to be. Having said that, this thread which is about the NIST's response hasn't uncovered anything that makes that response either incorrect or deliberately obfuscatory in nature. All you're actually citing here is that this girder would have been removed from the simulation (NIST term of art which I take to mean could no longer provide support) when it made a movement laterally of some specific amount, in this case 6.25 inches.


Again you cite a video which is not part of the report and represents a particular view of what happened as the girder walked off the seat. So, while I can see your point, that's not in any part of the report that the NIST produced which was my point. Remember that this started because it was claimed that the presence on an extra stiffening element meant that the girder couldn't have failed in the manner shown in the video. You've also stated that this is part of the report but it's not AFAIK. The leaves me baffled. To put it another way, you have developed your own view on what was meant by the failure of the girder described in the report and then you have produced a video showing that it's not likely - you say "impossible". The NIST response simply said the same thing. They said that the presence of the stiffener plate was irrelevant since no web failure of the girder occurred in the simulation that they ran.


The video depicts NIST´s narrative and simply shows what happens at 6.25" according to NIST.

Have you understood that the descriptions of the girder break it into 2 parts? _I_

Look again at girder on the cover of the video. The "web" is the vertical part that looks like an I, and everybody agrees that this did not fail.
NIST is saying that it was ok to omit the "web-stiffener" because the "I" did not fail, but it is omitting the fact that the stiffener plate ALSO stiffens the horizontal part of the shape _I_ which is supposed to have failed.

You are perhaps confused because the way NIST puts it, it sounds like the girder did not flex away because it says the "web" did not flex away.

Take a Look

If you simply look at figure 8-27B you can see that the NIST report posits no web crippling failure in the girder. It shows that the entire girder twisted and moved to the west. While I'd rather see something more detailed and sophisticated I think that this really settles the matter for me and should for you as well. The diagram shows that the web is intact but the vertical plate attaching it to the column has twisted allowing the entire girder to twist and the girder's end to be at an angle to horizontal. The twist is shown to be in the middle of the girder which is some 20 plus feet away from the stiffener plate that was left out. Apparently, the torque provided by the floor beam failures as well as the westward movement provided by their expansion and the general shifting of the girder itself by thermal expansion (must bow out/in as it's restrained at the columns from any east / west movement initially) is causing exactly what they said. The girder is rocked off (pg 353) is seat and there is no need to posit web failure. That's what their simulation showed.

I'm not going to try to address all the other points that have been raised about the report in this thread, just this one. It's simply not the case that the NIST has produced a report that depends upon the stiffener not being in place in order for a global collapse to occur. Thank you all for being so polite with someone who disagrees with you. There are many other forums where I am certain I wouldn't get the same treatment.


WRONG - you are confused..

What you have done is to throw away NIST´s final conclusion of "walking" due to thermal expansion and replaced it with an earlier ABANDONED theory of the girder "rocking off", which is now completely irrelevant.

ONCE AGAIN, to understand what actually happened according to NIST´s final theory, take a good look at that video and keep in mind previous comments. It should be easy now to spot the importance of the omitted stiffener plates.

Maybe this will help

This thread seems to be

This thread seems to be bogged down by some misunderstanding about what NIST claimed, and why the discovered stiffeners are so important in showing that their ‘initiating event’ was wrong.

I will retrace some steps.

NIST said in the final report that global collapse resulted from column 79 buckling as it lost lateral support over multiple floors. That loss of lateral support was said to be the result of a cascade of floors in that area. The cascade of floors was said to be initiated by a single girder spanning columns 79 to 44, on floor 13, walking off its seat at 79. (see item #11, NCSTAR 1-9 (Nov. 2008), page 611, PDF page 677)

‘Differential thermal expansion’ was said to break shear studs in the composite floor in that NE area, allowing five beams, framing into that girder, to expand as heat increased and push the girder laterally towards the West. The girder itself was said to have no shear studs which would then allow it to slide unhindered under the floor system, Westwards.

They said that the beams then reached a temperature of 600 C, and expanded by 5.5”, which is the maximum axial expansion possible because heat beyond 600 C causes steel to soften, and the beams would begin to sag, causing the span to decrease rather than increase.

NIST said that as the seat at column 79 was 11” wide, then that 5.5” lateral move would push the girder halfway across its 11” seat. As the full floor load then came onto the lower flange of the girder, without support from the vertical web, that flange would fail and fold upwards. The girder would then fall, taking the five beams, and the entire floor area in the NE with it.

The alternative ‘rock off’ theory required the five beams to buckle downwards instead of pushing axially, causing the girder to rotate ( downwards from the top) until it was pulled off its seat towards the East. Then the girder would fall, taking the floor area in the NE with it. More of this later. (see NCSTAR 1-9 Draft, Aug. 2008, page 353, PDF page 397)

Researchers discovered that in fact the seat at column 79 was 12” wide and questioned NIST’s assertion that the girder would fail at 5.5”, as the web was still supported by the seat and the flange was not under stress.

NIST admitted this error and said that a typo was the cause of the 11” / 12” mistake. They issued an ‘erratum’ document correcting that. However, in the same erratum they went on to admit another mistake. One of the transposition of two figures. They said that the 5.5” should be replaced by a figure of 6.25”, from another part of the report – and vice versa.

The effect was that the new move of 6.25”, across the new 12” wide seat would still allow the web to have moved over half way, and the flange would then still fail as before.

Researchers also had discovered that NIST omitted stiffener plates fitted to the end of the girder at the column 79 end. Those plates are clearly shown on all drawings. Such plates are normally specified by the designer to avoid web crippling when a girder is quite tall, and are fitted at a bearing point to prevent possible flex of the upright web at that point.

However, they also very effectively enable the flange to support far more load, as the plates also brace the flange to the web as well as brace the web to the flange. (see the video posted in this thread for more info on this). The effect would be to allow the girder to be pushed much further than even the new 6.25” and still not drop. If it doesn’t drop it then can’t allow a cascade of floors to occur and then column 79 to buckle. Without that column buckle the global collapse then can’t happen.

NIST was informed of the discovery of the stiffener plates and asked to explain why they had been omitted from all their own drawings when they were clearly shown on the construction drawings themselves. They were also asked to comment on the effect on their final report by this omission.

No reply was received, despite reminders, and that was the start of this thread in 911 Blogger.

However, since the thread began NIST eventually replied, admitting that they deliberately omitted the stiffener plates from drawings, and from all their analyses and calculations, because initial research showed that web crippling was not a factor and therefore including the stiffeners was not required.

Whilst this thread subject title is no longer valid following NISTs reply, the topic itself is far from over. The stiffeners make the whole ‘walk-off’ scenario impossible. And that makes the floor failure not possible. And then the cascade of floors can’t happen. Column 79 can’t buckle. Global collapse then cannot proceed.

Their alternative theory of ‘rock-off’ can also be overturned because other steel elements were also omitted from the report which would prevent beam buckling. But that aspect can be handled in another thread if required.

Meanwhile, another thread titled – ‘ NIST Replies to Stiffeners Inquiry’, here in 911Blogger, can take forward the discussions, to bring this important development to more people's attention.

8-27 Redux

The diagram (8-27) shows that no web crippling or flange failure at the seat as part of the collapse sequence that occurred in the simulation. Rather the top clip appears to be failing, allowing the girder to rotate at the seat. I do understand that you think that the stiffener plates would have helped to stiffen the girder in a general sense but I don't see how this helps your argument - the actual twist is happening far away from the seat but the torque gets transmitted to the seat/column/girder connections.

That's why NIST responded to you(?) as they did - the plates were omitted because there was no crippling of the web or failure of the flange at the seat. It would be interesting to talk about your other points, which I have not researched very well to date, some other time. As far as this one is concerned, I think that I understand your position, but NIST is telling you that the failure model which came out of the simulations made those stiffeners moot. You may want to try a different tack and do some calculations and demonstrate that the girder can't rotate in the seat in the manner shown in the diagram by doing some math (beyond me at this stage in life).

The video shows a linear lateral movement of the girder westward, this is also different from what was depicted in the report, FYI which shows a bowing as well as the aforementioned twisting.


Hi Woody

My question to you is what would cause the girder to fail when NIST supposed it was pushed 5.5" (revised to 6.25). I believe they said that the flange was unable to support the load. So what would happen to the supposed overloaded flange at that point?
Also, NIST claimed that the girder may have been 'rocked off' to the east, as well as claiming the push to the west. I think that is perhaps where some of the confusion on this issue is coming from. As for the video, I made that, and I stand by what it says 100%. I'd be happy to take the time to discuss the whole issue of the stiffeners with you here or elsewhere, skype for example, as I don't think it would take long to clear up the confusion. Feel free to message me to this end.


If I'm misreading the report, please just show me where I'm going wrong. I would like to do a skype with you. Is there a way to PM on this site?


With all due respect, I am asking a straightforward question above. If the flange did not buckle, what caused the girder to fail, considering it had not been pushed beyond the seat plate in NISTs analysis. Are you saying that it failed without the lower flange distorting?

Who are you kidding, "Woody"?

You refuse to acknowledge that you are confusing an alternative "rock off" hypothesis with the "walk-off due to heat expansion" final hypothesis which NIST chose to represent the initiation for global collapse, and you are pretending that you have not been informed about this here: http://911blogger.com/news/2013-09-25/60-days-nist-refuses-reply#comment-260362

NIST does not base anything on 8-27, it was an exercise to demonstrate an alternative theory that is irrelevant to the final conclusion which is the "walk-off" due to heat expansion.

If you cannot acknowledge your mistake you would appear to be doing this on purpose, which would mean your only intention is to waste our time aka trolling, in which case you will be ignored and perhaps even banned from the forum.


If I'm mistaken/confused then so be it. Can you present some specific details about my mistake?

I read the report and it seemed to me that they did, in chapter 8, detail what they think the leading collapse hypothesis was. This was tested that through an initial simulation (16 story Ansys) the results of which are presented in Section 8.8. That simulation showed no flange failures and no web crippling. What it showed in 8-27 is a twist and failure of the top clip. If you can show me where in the report they say there was a different "final hypothesis" which contradicts Section 8.8 then I withdraw my comment about 8-27. I had looked at 8-27 a few times in the past, but actually just realized what it was intended to present yesterday.

For example, we can look at the details in the second volume of the report where the term "walk off" in used Pg 505. It's always been my reading that the term walk off is NOT equivalent to a simple lateral motion for the girder and should be understood in the context of 8-27. This is why I think that the 8-27 diagram shows the most likely failure mechanism if the girder can do a simple linear walk off sufficient to move it off the seat. Rather the end is subject to a large torque because of the twist in the middle shown in the diagram - the top clip fails. Otherwise you are positing that the only forces acting on that girder are due purely to thermal expansion, that can't be the case if the beams are shown to fall away after some period of time.

That's how I read the report and it's inconsistent with the requirement for there to be any flange or web failure. If you stick to the premise that the report requires a simple linear walk off of the girder then, absent a full analysis of the forces acting on the girder you have a case.



Chapter 8 scenario 8-27 is not the leading hypothesis - it was entertained but then ABANDONED for another hypothesis.

The "detail" of your mistake is that you are confusing and conflating 2 theories that contradict each other so much that they are in direct opposition to each other: 1. the old abandoned theory of beam shortening and girder rocking off(represented by 8-27) VS
2. the chosen final theory of heat expansion leading to that same beam extending and girder walking off in the opposite direction!

Take another look: http://imageshack.com/a/img31/4332/6obx.jpg

By conflating the 2 theories you are resorting to gibberish because you are trying to combine theories that say the girder was pushed to the west and fell off the west side.....AND that it was pulled to the east and fell off the east side!!!!

And most importantly of all, even though this MAY make sense in your head, even NIST does not try this kind of BS: 8-27 is NOT part of the final hypothesis and has absolutely nothing to do with it! If you want to keep this up I suggest you move to the JREF forum and stay there!


First, I'd be interested in seeing some documentation rather than just an assertion that 8-217 represents an "abandoned theory". Second, regardless of the relevance, it is illustrative of two points. First that the report doesn't preclude rotation as part of the walkoff mechanism. That's your assumption and it remains difficult, at least for me, to work out why they would include a large color graphic showing something that they had abandoned and wasn't part of the collapse mechanism. Even more telling is my original point that since the report shows neither web crippling nor flange failure it is not at odds with what the NIST said to you. It is simply at odds with the video that you have produced. Finally the argument that the stiffeners would prevent significant actual rotation/distortion of the girder is not backed up by anything. I'm happy to concede that they must impart some stiffening but whether that's enough to prevent rotation and walk off is just a guess on your part.

Last time I posted at JREF they accused me of being a "truther" so I don't go there any more.



To obtain a 5.5+ inch walk-off of the girder to the west through beam expansion you would have to accept ALL of the following.

To obtain a buckling of the eastern beams you would have to accept ALL of the following.

1. 4 hour fire
2. Shear studs in the composite floor system break throughout the region allowing the beams to expand
3. No stiffeners on the C79 end of the girder
4. No beam stubs on the G3005 north beam

The fuel load supports 30 minutes. NIST's model used 240 minutes.

Start there.

new theory

Woody has abandoned the the heat expansion/walk off theory, the question is whether or not he knows it.

We'll See I Guess

I think that I read the report differently from you. I think that a simple linear walk off of the girder from the seat in not what the report says. I think that there are other forces acting on that girder and that we can see them in 8-27. I don't mind being wrong, but I can't find anything in the report that implies that combining the two elements of linear and rotational movement contradicts the findings.

Interesting Comments

If you want to start another thread on this, fine. I'd like to stick to this one for now as it relates to the video and the NIST response and that's really enough for my brain.


You obviously "stick" to your own preferences no matter what - I don´t know if this is due to limits of your brain or if you are just being stubborn or perhaps just "trolling" - so it is fair to say that any further attempts to reason with you would be a waste of my time.

Please refer to following post if you are genuinely confused: http://911blogger.com/news/2013-09-25/60-days-nist-refuses-reply#comment-260386


I read your post, it's just an assertion. Do you have anything to back it up? When I read the post, I figured that I was probably wrong, actually - it seemed too obvious that the top clip fails and the beam rotates so as to add an imbalance which would simply be an example of something more complicated going on that a simple horizontal walk off. That's really at the center of the report, I think - that you need to take into consideration all of the forces which are on that girder to understand why it walks off the seat - it''s just not as simple as you are choosing to make it. Your arguments appear to be directed at the proverbial straw man.

Rock Off Can't Happen

Please take the time to study the conditions related to the Rock Off theory.

The beam stubs on G3005 will prevent this from happening. Fires will not last long enough to heat up the beams to the point of buckling. Shear studs will not fail.

This never happened in office fires before. Why are you so willing to accept it for WTC7 on 9/11?

NIST wants it both ways. They want beam sagging/buckling to support a rock-off and they want beam expansion to support walk-off.

That´s chapter 8, but....

But chapter 8 is not the whole report and there is actually a discreet disclaimer right before graphs 8-26 and 8-27:

"This analysis demonstrated possible failure mechanisms which were used to develop the leading collapse hypothesis further"

"Possible" but not actually the mechanism chosen in the end, some details did change.

Later chapters (11, 12, 13) where NIST describes the same initiating failure sequence, this contradiction has been resolved: the "rock off" theory has been omitted from all chapters after 8 and replaced with a new cleaner mode of failure...the final conclusion only mentions thermal expansion and walking off at 6.25", as has been made abundantly clear with several citations from these chapters.

NIST should have revised chapter 8 to make this clear, so maybe it is understandable that people get confused. HOWEVER, Woody should have understood the error after having it pointed out to him several times, in several different ways...but somehow I doubt he ever will.

Again, who are you kidding, "Woody"?

My back-up?

You know, the REST OF THE REPORT - try going beyond chapter 8! See response to kawika above